Giray Unlu

Machine Learning Engineer at Innovai ITS Solutions Inc

Adapted image classification models by fine-tuning and training from scratch to detect various types of products and apply in product-based algorithms. Optimized object localization models to detect specific signals within images, achieving 96% accuracy and a high F1 score. Applied computer vision algorithms for image and color analysis to convert signals into binary arrays over time. Scaled and classified time series data for action recognition cases by creating data using computer vision algorithms and models, including YOLOv7. Applied ML algorithms to field-focused tabular data, achieving ~94% accuracy and a high F1 score.

Technical Team Lead at METU AI Society

Led deep learning projects, including reproducing code from well-known research articles and implementing state-of-the-art models. Managed problem-oriented computer vision and data science tasks for public datasets; oversaw two projects per semester. Delivered lectures on deep learning and machine learning to university audiences and mentored new members in the field.

Machine Learning Engineer at Innova ITS Solutions Inc

Adapted image classification models by fine-tuning and training from scratch; optimized object localization models; scaled and classified time-series data for action recognition; applied ML to field data; developed and deployed data-driven ML pipelines; engineered prompts and tool use for advanced LLM applications.

LLM Engineer at Salesport Technologies GmbH

Developed and maintained Retrieval-Augmented Generation (RAG) pipelines using language models with vector databases and cloud storage for improved data retrieval and accuracy. Designed and deployed autonomous LLM agent workflows with integrated reasoning and tool use. Developed and optimized web scraping tools for data collection and analysis in ML applications. Engineered prompts and utilized advanced LLM capabilities, including structured outputs and function calling, to optimize model performance in real-world applications.

Machine Learning Engineer at Bueno Art Inc

Fine-tuned and built various Stable Diffusion models for different styles and objects; implemented scripts for multiple Stable Diffusion frameworks; designed and deployed AWS-based pipelines for modular product-level models (3-100 MB).

LLM Engineer at Sport Technologies GmbH

Developed retrieval-augmented generation pipelines using LangChain; integrated large language models with vector databases and cloud storage; designed autonomous LLM agent workflows and tool usage; built data-collection and analysis pipelines; crafted prompts and utilized function calling to optimize model performance.

Machine Learning Engineer at Innova IT Solutions Inc

Adapted image classification models by fine-tuning and training from scratch to detect various types of products; optimized object localization models (e.g., YOLOv7) for signals within images; scaled and classified time-series data for action recognition; mentored new members in METU AI Society.